Loudness Standards in Film, TV & Games

notlikethisYou may have heard of the “loudness wars” in the context of music before, but it also exists in film and television productions. Just like mix engineers of music abused brick-wall limiting to reduce dynamic range to fool loudness standards based strictly on peaks and RMS, so too did mix engineers for video-based productions. This comes from the desire to be the most attention-grabbing experience, however it often bastardised the depth of the soundtrack, leading to static aural experiences. You may once have looked like the guy on the right when a TV show went to an advertisement that was far too loud due to compression.

In the case of film and television the ITU (International Telecommunication Union) and EBU (European Broadcasting Union)  have devised new standards that combat this exploitative practice, which does a better job of standardising the average volume of productions while retaining dynamic range.

Loudness Measurements

Loudness Units (LU)

Loudness Units, unlike LUFS (which are directly proportionate to dbFS), is a relative measurement that is used to describe the difference in Loudness Range, as well as True Peaks and the satisfactory deviation from Programme/Integrated Loudness in the context of loudness metering.

Loudness Range (LRA)

The Loudness range is the descriptor of the difference between the loudest and quietest sounds over the course of a piece of media, which ignores the loudest 5% and quietest 10% to avoid the influence of such things as gunshots and fade-outs of music as they are not representative of the main body of audio (“Loudness Explained”, 2012), and each loudness standard has their own acceptable LRA to abide by. This is not to be confused with dynamic range, which is the difference between RMS level and peak level.

iZotope Insight

iZotope Insight, one of the various tools available to measure loudness

LKFS/LUFS

LKFS, or Loudness, to K-weighting, relative to Full Scale, was the first introduction of a way to measure the perceived average volume of a piece of media over it’s duration. It came about when the ITU wanted to address complaints from consumers that the difference in their perceived level of commercials compared to television programming was vastly different. This was due to many programmes having a much larger dynamic range than commercials, which were heavily compressed. This was because of the previous standards’ normalisation of audio being based on peak levels.

LUFS, or Loudness Units (to K-weighting) relative to digital Full Scale, was introduced when the EBU noticed a problem with this methodology of measurement. Forms of media could insert large durations of quietness, if not silence, to offset this measurement. The solution that the EBU came up with was to introduce a gate to ignore audio 10 LU (Loudness Units, a relative measurement) below the absolute loudness level in order to combat this practice, and to provide a more accurate measurement of the listener’s perception of loudness. At this point EBU introduced Loudness Units and Loudness Range (LRA) as additional measurements as well. In 2011, ITU released BS.1770-2, at which point both ITU and EBU have agreed to use a gate of -8 LU (Taylor, 2012). Since this event, there has been no difference between the measurements. Now, LUFS have well and truly become the standard measurement unit for the average loudness of a mix.

Standards

Before I jump into each specific standard, it is important to note that there are two methods of metering loudness that generally practiced and adapted across each standard, one is based on the ITU BS.1770-3 ideology that relates to Programme Loudness (LKFS) and True Peaks (the estimated peaks between samples post-ADDA conversion), and the EBU R128 ideology that relates to monitoring Integrated Loudness (LUFS, the same as programme loudness), Momentary Loudness (loudness over a window of .4 seconds), and Short-Term Loudness (loudness over a window of 3 seconds)

OP59 – Australia

The operational practice 59 (OP59) was introduced in 2012 and replaced the old OP48 standard which relied on peak metering. The OP59 standard references the ITU BS.1770-3 ideology of loudness metering and dictates that pieces of media adhere to a Programme Loudness of -24 LKFS within a deviation of +/- 1 LU . It also specifies that True Peak should not meter above -2dBTP (“Free TV Australia Operational Practice OP – 59”, 2016).

ATSC A/85 – America

The American standard for Loudness was introduced in 2009 by the Advanced Television Systems Committee. It also follows the ITU BS.1770 ideology of Programme Loudness and True Peak metering, though prior to 2013, it only took Programme Loudness into consideration for commercials, and instead uses an “anchor” system to regulate loudness in regular programming (“Broadcast Standards”, 2013). You can learn more about the anchor system in this document from the Audio Engineering Society. After 2013, the A/85 was updated so that all programming should be required to meter Programme Loudness. The Target Loudness under this standard is -24 LKFS with an acceptable deviation of +/- 2dB, and True Peak should not exceed -2 dBTP (“ATSC Recommended Practice”, 2013).

TR-B32 – Japan

The Japanese broadcast standard TR-B32 is based on an earlier version of the ITU methodology of loudness metering, ITU BS.1770-2, which employs a different gating system to the ITU BS.1770-3 metering methodology, and sets a Target Loudness of -24 LUFS.

Loudness Standards for Games

While there has been no specific standard set for games, a recommendation has been made by the Game Audio Network Guild IESD (Interactive Entertainment Sound Developers) for Sony, Microsoft and Nintendo console titles (GANG IESD, 2015). This recommendation is based on Sony’s internal ASWG-R001 standard from 2013, which uses ITU-R BS.1770-3 metering (Taylor, 2013).

Due to the unknown quantity or content of games, the target LUFS measurement is a little more flexible, and a different procedure is undertaken to acquire the measurement. The GANG IESD recommendation specifies that as long a segment of gameplay should be measured as is practical, however, if this is not practical, multiple chunks of gameplay should be measured, which should be a minimum of 30 minutes each and loudness measurements should be averaged between them. Each of these chunks of gameplay should also be representative of the game as a whole rather than an especially quiet segment or an especially loud segment.

The GANG IESD recommendation for console games strives for a loudness measurement of -23 LUFS, or -18 LUFS in the case of mobile or handheld games. Both loudness measurements have a toleration of +/- 2 LU, and the GANG IESD recommends an LRA of 10-15 LU for mobile or handheld games. The True Peak limitation in this recommendation is -1dBTP.

References

ATSC Recommended Practice: Techniques for Establishing and Maintaining Audio Loudness for Digital Television. (2013). Advanced Television Systems Committee. Retrieved from https://www.atsc.org/wp-content/uploads/2015/03/Techniques-for-establishing-and-maintaining-audio-loudness.pdf

Broadcast Standards. (2013). Retrieved from http://www.tcelectronic.com/loudness/broadcast-standards/

EBU (2014). Loudness Normalisation  and Permitted Maximum Level of Audio Signals. EBU. Retrieved from https://tech.ebu.ch/docs/tech/tech3341.pdf

FREE TV AUSTRALIA OPERATIONAL PRACTICE OP – 59: Measurement and Management of Loudness in Soundtracks for Television Broadcasting. (2016). Free TV Australia. Retrieved from http://www.freetv.com.au/media/Engineering/Free_TV_OP59_Measurement_and_Managemnt_of_Loudness_for_TV_Broadcasting_Issue_3_July_2016.pdf

GANG IESD (2015). GANG IESD MIX RECOMMENDATIONS. Game Audio Network Guild. Retrieved from http://www.audiogang.org/wp-content/uploads/2015/04/IESD-Mix-Ref-Levels-v03.02.pdf

Loudness Explained. (2012). Retrieved from http://www.tcelectronic.com/loudness/loudness-explained/

Lund, T. & Skovenborg, E. (2014). Loudness vs. Speech Normalization in Film and Drama for Broadcast. Audio Engineering Society. Retrieved from http://www.aes.org/technical/documentDownloads.cfm?docID=521

Taylor, G. (2012, February 20). What’s the difference between LKFS and LUFS?. [Web log
post]. Retrieved from http://gameaudionoise.blogspot.com.au/2012/02/whats-difference-between-lkfs-and-lufs.html

Taylor, G. (2013, April 12). GDC 2013: Garry Taylor – “Audio Bootcamp: Loudness and How to Measure It” [Recorded Lecture]. Retrieved from https://archive.org/details/GDC2013Taylor

Advertisements